GTX 480m officially confirmed, coincides with E3/Alienware refresh

You mean the roadmap rumor I posted a couple of days ago? Even if that unofficial roadmap isn't make-believe Blackcomb isn't marked for launch until 2011.

 

just about every major notebook manufacture has had a 5870 in their laptop for months. AW isn't already behind them? For the year or so of hardware updates so far, I have never seen the M17x *ahead of the game*(until now with CF 5870s) They've been able to catch up most of the time but never beat most of the big names (sans clevo and eurocom) *totally* before the punch. Even with CF 4870s, Eurocom had been selling SLI GTX285s, which are just as good if not a smige better when you consider hybrid graphics, and physix.

lol WHAT? After the 5870's being avaliable for less than a month before the 480M's were officially announced?they took the time for a core clock speed increase and API update card like the 5870, so why not something like a 480M?

I Apologize but the wording still has me confused, as you suggesting that because they mention N11e(GTS360M) fixes that means 480M too? Well, first, it's not under the A05 Bios, according to the release notes, the only two things added in A05 was VT and a Broadway XT fix. Where it is mentioned is in A01 and A03, which are dated up to 03/01/2010. How could they have added support to a card that Nvidia was 3 months away from even announcing the existence of(and then called it a card to which it has 0 relation to)? IMO, they were considering a Low cost/ Low powered GPU (like the GTS360) to sell to people that wanted a bit more mobility(since we don't have hybrid graphics any longer) or a system at a cheaper price but just never followed through. You're talking about a clock speed updated GTS 260M (which is slower than the slowest card ever in the M17x, the GTX260M) compared to a fermi-hybrid. I just don't see how it's possible considering the amount of time they took for something(the 5870) that has *far* more similarities to an existing card in the system (the 4870)

 

Though this test is inherently unfair in terms of GPU comparison, the GTX 480M should and will be the faster of the two. However, from the looks of it, it wouldn't be my much at all. A 5870 with a 920XM + newer drivers scores 7800 ish for Vantage GPU at performance and 5000+ at high settings. Also, looking at this test between the desktop versions, it looks like the roles are reversed in terms of who makes better/more improved drivers.

Some of you remember me saying that it's 8500 vantage GPU @ stock settings or bust for this card, yet it's barely touching 8000 with a desktop i7 980X, making it saturated with 6 cores at 3.33ghz. It might do a lot better in other games, but I was really expecting it to have more raw power so that it can beat ATi even at ATi friendly games (hence the whole 8500 Vantage or BUST). A Bust for now, maybe drivers would help.

Now in SLi, this might shine a bit seeing as Fermi in SLi scales much better than crossfire in most cases. And if there is 2gb of video memory, it would be even better as there would be more room for memory when mirrored.

 

Those tests are worthless. It shows NOTHING except a 480 is a monster card when put together with an insane desktop rig.

For having almost twice the wattage the card is actually kinda pathetic. I was hoping for something much more spectacular for that kind of power draw to make a really good single card system viable again, but looking over even these completely skewed tests thats just not going to be the case.

I wonder why they just didnt put the ATI cards in the same rig and test that way... false advertisement at its best

 

Where has nvidia released power consumption figures for the 480M? TDP is not power draw. I'm sure I don't have to explain to you that watt's are measured in more than one way. Although I agree, little can be extracted from tests with vastly unequal components.

Also, if you *really* want to snob about double the TDP, I'm sure the other 40 dozen people that went on in the 5870 thread about the inaccuracy of such numbers could chime in. Even AMD suggest the 5870 is now a 50W TDP chip, even though clearly when the GDDR5 is taken into account, it is not. Previously it was 60W, and I'm not sure where that page that said that went. The only indication of the old 60W TDP figure is from the Crossfire 5870's TDP of 120W @ notebookcheck's list of Mobile graphics cards. I would not be the least bit surprised if it were actually in the 75-80W range *with* the GDDR5.
http://www.amd.com/US/PRODUCTS/NOTEBOOK/GRAPHICS/ATI-MOBILITY-HD-5800/Pages/hd-5870-specs.aspx

 

Lozz you are on a roll today... Though I understand most of the points you are trying to make you forget to mention a few things:

1. Implementation takes months and I am sure AW has been working with the 480M for weeks if not months.
2. Implementation of a new card will only happen if it makes economical sense and that would be only the case if AW was sure to make money on it therefore giving plenty of time for customers to buy. Of course that might not be true if the cards were new versions of the ATIs but they know some people will still choose the 5870XF even if the 480SLI are an option.
3. Concerning your explanation on the TDP I totally agree but nonetheless you cannot deny that the 480M has a thermal design unheard of in the mobile world and it's power consumption must be huge also.

 

Strangely enough I do realize that.

And yet still i blast it for being grossly inefficient.

I had never heard the 5870's were to be 50w chip i thought they were 65w TDP. I never really researched it though as i'm not really interested in getting the cards

Thats pretty crazy though (the referenced website).
I mean if they designed a much better heat sink (solid copper that actually contacted the memory and chip) then i "might" see it as a 50w chip. Thats either an underestimate of the needed cooling or GDDR5 can handle its crazy heat alot better than GDDR3 could.

But as to being Snobish... i dont think i am. I've ran nVidia in almost everything i've ever owned (not counting integrated stuff) and i generally prefer them, but this card is not a very good solution for mobile platforms. The trends is supposed to be more efficent process and better capabilities, not "throw enough muscle at it until it works"

 

Did they turn off the PhysX to get those nVidia 480m score?!?

Look at the CPU scores, G73 may ship with i7-720, not the strongest mobile quad - but it's 400%+ weaker in vantage neither. i980 may have 50% more core, higher frequency, but the score still shouldn't be that different. Closely examining those nVidia ones, which seem to be ahead by miles, It lead me to believe that whoever has done the benchmark, allowed 480m to run with PhysX on.

Though while 480m is definitely more powerful in DX11 games, the difference is less so in DX10, more over, those DX11 are hardly playable at times. It's only the first wave of DX11 GPU though.

 

I disagree, otherwise they seem to like to take double the time other OEM's do in testing. Their past history with nearly every other graphics card says otherwise. I'll reiterate that ASUS, and many other OEM's had 5870's in their laptops months before Dell released a 5870 M17x.

you're right, I don't deny it. It's ridiculous, but nothing Dell hasn't handled before.

orginally,
4870=65W
5870=60W b/c of 55nm reduction
but now on AMD's site it's magically 50W. I'm not sure what went on in the between time. I *do* know that the 5870's dell use are low voltage boards (1.05V vs 1.2V) so it may be that AMD was able to refine their 55nm process a bit or simply that they can run that low consistently and adjusted the TDP to reflect this.

perhaps, however I suspect that it will not be as big a deal as people make it out to be(or rather something that they will not consider a downfall), unless we have another manufacturing process 9800 disaster. Computers are becoming more power hungry and there's nothing we can do about it. If people want high performance machines increased TDP and power draw are *going* to be an increasing side-effect. Dell successfully implemented the 100W quadro 3800M in the Precision M6500, and the 480M gives us almost triple the cuda cores, 256bit GDDR5(which is *HUGE* needless to say) and directx 11 within the same TDP, I don't see why the M17x should be any different. -Or- why that makes it so grossly inefficient, from that perspective we're getting a lot more for a 100W

 

So how much power (watts) can the m15x handle max from its graphics card. Because I dount the 480m will work with the m15x.

M17x if its 100w without a doubt as according to notebookcheck, the 5870's are 120w.

 

Yes, 5870XF is 120 watts but that is for two cards. You are talking about 480M SLI with 200 watts...

We all know a single 480M should be available soon enough just like they offered the 5870 way before the 5870XF.

 

only dell has the capability to answer that question(purely from a 'lets shove X in here and see if it works perspective). However, plenty of people run the early GDDR5 5870's which were ratted @ 60W in their m15x without issue. They're able to shove a declocked 5870 (the 5850) in there and sell it sucessfully, so I'm sure the M15x's cooling capability is close to or as good as the M17x's. As joe said above, the 120W figure is from two cards in crossfire mode. They're 60W individually.

 

The numbers we get from AMD and Nvidia are nonsense, so don't even waste your time on trying to understand.
MR5870's TDP is for the chip alone. Plus, what tools were used in measuring? What calibration levels? Etc...
The only real life test is to take a single system and test it with 5870 and then 480M and measure the power draw and heat with the same tools. There was a similar test on Clevo's boards 5870 vs 285M and they appeared to draw almost the same power and produce the same heat. So it looks like the "real" TDP of the 5870 is closer to 75W if you can assume that 285M has a 75W TDP, lol.

 

essentially, you just stated what I was attempting to say, although indirectly. so i agree.

 

Yep, totally agree.

 

Well I'd like to see a video where they cram
1. a gtx 480m and see how a m15x handles (with possible burnouts and failures etc..)
2. Gtx 480m scores in EQUAL machines versus the 5870m. Not some stupid 920xm machine with a 480m versus a 720QM machine with a 5870m. We need to use the ceteris paribus rule in this case (latin for all things remain constant) obviously apart from the graphic card.
3. See how the CURRENT M17xr2 fare with 480m SLI. Would it have the same effect as the m15x with a 480m or will it do better/worse.

Interesting things for people to look out for if someone can experiment with this (obviously those with lots of computers to spare lol).

 

We went over this so many times ). We've concluded that it is around 65-70W for the whole card right?

 

Yeah, I think so.

 

Certainly the TDP is an inaccurate measure but I remember seeing benchmarks the other day comparing desktop GPU which stated that nvidia's GTX465 reached about 10°C higher than ATI's 5850 both in idle and load with otherwise exactly the same rig... while performing worse actually. I'll let you draw your own conclusions as to what is to be feared / expected from mobile versions. Nvidia has definitely not come up with an efficient solution when taking in account the performance / heat ratio, imo.

 

There is no doubt the cards are going to run warm/hot.... Also no doubt we will hear that 94+ degrees while gaming is perfectly fine etc ...

 

Especially for the memory

 

Well my stove is starting to get a little worn down for cooking my eggs and I do intend on buying a new laptop, PERFECT!.....

 

I know the card is going to run hot/warm when i saw a reseller say something to the effect that the card runs at acceptable temp levels etc...

Any time i read comments like that i know the thing will be a toaster and also why we dont see any temps yet even though certain resellers etc have these machines in hand? i think the only thing under NDA are the temps lol...

Still very excited to see the outcome of the card regardless of any shortcomings ...

 

Was poking fun Quad, I think the cards will be managed properly by the ever solid Clevo, but I mean still I can make doubleuse of it to fry my eggs, bonus for sure!

 

It's only a matter of time. Give it a little bit, this is June and they *should* be released within 2-3 weeks. I'd like to see how they perform in Alienware's SHOULD they decide to incorporate a single 480m.

 

Yeah i was typing my comment before i even saw yours so mine was not directed at yours ...

As for frying your Eggs yes you will certainly be doing just that ...

 

Meanwhile look at the M7820 (5870) temps in a 8740w.
And here during BC2!

 

Showoff!

Those are great temps; that's why im sticking with 5870's regardless of nividias hullabaloo

 

Those temps look very good and similar to the W870CU i had... I think my max in furmark was 80...

 

People with 5870CF in M17X-R2 report 95C+.
Suddenly I doubt the superior and famous cooling of the AW's.

 

The highest mine ever hit was 93c which was in furmark

 

Well it took some work with the heatsink/paste etc to get my temps to where they were on the Clevo but overall the Clevo heatsink is a beast of copper and nothing like the Alienware one...

 

I don't think anyone is confusing it to be superior, adequate and smart maybe. Also, name another Crossfire 5870 laptop that runs cooler, Whoops, you can't. Unless you're expecting a Dual GPU laptop to run it's cards cooler than a single GPU laptop (which is absurd) then there's no argument here. The very nature of the comparison is flawd since a dual GPU laptop is going to have to conserve real-estate more to be able to run two smaller heatsinks versus the single GPU having the ability to be more liberal with it's design and heatsink size.

 

Plus it's the memory modules and controller hitting 90+, not the core. So it's more of a fault that the cards were engineered on a 128 bit bus with gddr5 than a failing of AW's cooling design. I'll take slightly hotter temps of an AW over a cooler running Clevo plastic lunchbox any day of the week. BTW Aikimox, in actual games at stock speed, my temps (including memory) never exceed 80-83C. It's only when running something extreme like furmark xtreme benchmarking mode that it hits 90+ for the memory.

 

Is the X7200 plastic? It looks sort of metalish in the computex photos. Either way, I can never own a Clevo because I NEED a backlit keyboard. Asus, HP, or Alienware for me.

 

You sound like an AW fanboy now, lol.
The mere fact that I own AW systems and like them very much, doesn't justify the blindness. Look into the X8100 and D90xC forums to get the answer to the temps questions.
I can only partially agree to the 2 smaller HS's vs 1 huge, since there's a lot of room for improvement for the current M17X cooling.

You know that the memory is more sensitive to higher temps, right? Saying "My core runs cool, only mem is at 85-90C" is ignoring the fact that the card will die just as fast as if having 90C on both core and mem.
And you know what, after looking at few different cooling systems with 5870 and their temps, I think AW did a poor job on cooling the memory. It could be kept on par with the core, IMHO.
I agree regarding the plastic boxes (another fact, I'm here with you on the same board and not there with a lunchbox full of copper.) But it doesn't mean that I will stay oblivious to the flaws, be they in cooling, plastic palmrest, crappy touch keys, etc.

Flame wars are for fanboys,
Just want AW to evolve simply and especially because I like their products and consider them one of the best. This might change, if they continue to neglect the flaws.

 

First of all, I'm all for metal laptops. I do foresee multiple issue with metal palmrests though. The first one is that less heat will be transferred to your palms though plastic than through metal, but internal design changes may be able to regulate how hot the palmrest gets. The other issue with metal palmrests is sweat. I'm not sure how much sweat gets on your, or anyone else's palmrest after a gaming session, but I know that I literally corroded by M1530's palmrest with how much I sweat on it. A friend of mine has done the exact same thing from gaming so much on his MBP. Let me see if I have any pictures of the corrosion I'm talking about.

Edit: I can't find the picture, but I can describe it in more detail. It appears to be something comprised of black speckles and dots that is securely stuck to the palmrest. After failing to remove it, I inspect it closely. It's actually concave and not convex.

 

After spending roughly 200-300 hours gaming on my R2, palmrest became very wavy and it gives in with a cracking sound. Same happened to my R1.

Not every system with a metal palmrest has to experience overheating or corrosion issues. You know, they just have to use a high quality finish to make it last longer. Look at HP8730w for example. People game on it with no issues, AFAIK. And it's thin

 

Dude you know I criticize Dell/AW when they deserve it. However, how do you expect them to have designed the M17x around the 5870's insanely running hot memory? The system was designed over a year ago for much tamer hardware. The fact that they have been able to adapt it to run within operating tolerances of the 5870m crossfire speaks volumes about the quality of their product and cooling design. The next Dell/AW system will just improve on that and blow us all away. Oh and about the X8100 and D900, the X8100 was designed and released much later than the M17x. The D900 is designed to cool a desktop CPU, not exactly an apt comparison there. I don't expect to see failing 5870 memory modules within the next few years because of the temps they operate at. It's within their designed specifications. The MTBF shouldn't change much at all as long as the parts weren't designed with flaws from the get go.

 

They could at least redesign/mod the heatsinks focusing on the memory and provide better quality pads to begin with. I'm getting my elitebook in a few weeks and will compare the heatsinks and maybe we can find some ideas. They managed to keep mem and core temps practically on the same level.

But I can understand AW/Dell' reasons , they had to act fast and still managed to do a fairly decent job. And you're right they will most likely do a better job for the gen.3.

 

Are the heatsinks any different for the 5870s compared to the 4870s Joker or are they same ones carried over ?...

Also as far as plastic lunchboxes(lol) yeah i cant say im a huge fan of Clevos choice of exterior parts either but internally its a whole different story...

If AW could incorporate all that copper like Clevo does then maybe they could also bring the fan noise down to a reasonable level ... The 8760 i had never got loud at all so it was bliss when gaming but i wish i could say the same about the M17x .... Id love to order another R2 with 5870s right now but that noise i know would make me hate the machine instantly all over again so i wont even waste the time ....

 

I agree about the fan noise, it's a bit much and I'm guessing they'll do something about it with the next M17x or M18x (should there be one) release. That said, the noise doesn't bother me if I"m gaming cuz I wear headphones most of the time. During regular web surfing or other non-gaming tasks, it runs fairly quiet.

 

There wouldn't be much copper inside if Clevo went with a metal exterior They'd be forced to switch to aluminum alloys, otherwise the entire case would turn into one huge BBQ grill. Copper conducts heat better than aluminum but dissipates slower and it would mean more heat trapped in the metal case.
I wish to prove wrong, but (IMHO) there's no way to have the power of X7200 with a metal exterior.

 

Yeah well thankfully the X7200 is not on my list of "Must have" machines....

From a technical standpoint it looks amazing but aesthetically its not my cup of tea ... From the top down i thought it looked fine and basically just like the 8760 and you can see they took the design from that machine and then strapped on a basement underneath it .... Anxious to hear the numbers it pulls down though with 2x480s ...

 

How do you think they power those dual 480s? They've got a elves working in there.

 

Has anyone compared the size of the fans used by Clevo and AW? I bet Clevo uses larger ones to keep the noise level lower. Probably the new AW laptop will use larger fans to have efficient cooling with lower RPM thus noise levels.

 

Haha i knew there was a reason the bottom was so big lol... I was thinking smurfs ...

 

Not sure about the newer Clevo's but my M570RU is loud at full blast. Sounds like a hair dryer.

 

The Clevo x7200 dimensions are 16.5" x 11.26" x 2.4". What is the M17x's anyone?

 

The new one was not much louder under load then it was during idle ... I know the old ones were loud but nothing holds a candle to the M17x thats for sure... Its like running 4 hairdryers at once...